Method for Retrieving a Tumor Contour of an Image Processing System

1. A method of retrieving a tumor contour implemented in an image processing system that includes a memory storing a grayscale image, a display screen, an input device and a processor performing method steps, comprising: receiving an input tumor contour of the grayscale image; defining a tumor contour annular region and a plurality of reference segments of the grayscale image, wherein the input tumor contour is in the tumor contour annular region, and each of the plurality of reference segments is across the tumor contour annular region and includes a plurality of measured points; retrieving a tumor contour suggestion point on each of the plurality of reference segments; and linking all the tumor contour suggestion points on all of the reference segments, for forming the tumor contour, wherein each of the plurality of reference segments includes a plurality of local segments and each of the plurality of local segments includes a plurality of moving horizons; wherein each of the plurality of local segments consists of one of the plurality of measured points, at least one measured point before and at least one measured point after the one of the plurality of measured points, and each of the plurality of moving horizons consists of one of the plurality of measured points in the corresponding local segment and at least one measured point after the one of the plurality of measured points; and wherein the tumor contour suggestion point is retrieved by a moving variance retrieving method comprising the steps of: calculating a gradient variance of the grayscale image in the plurality of local segments and an average gradient variance of the grayscale image in the plurality of moving horizons; calculating the ratio of the gradient variance of the grayscale image in the plurality of local segments to the average gradient variance of the grayscale image in the plurality of moving horizons; and retrieving the measured point corresponding to the maximum ratio as the tumor contour suggestion point.

2. A method of retrieving a tumor contour implemented in an image processing system that includes a memory storing a grayscale image, a display screen, an input device and a processor performing method steps, comprising: receiving an input tumor contour of the grayscale image; defining a tumor contour annular region and a plurality of reference segments of the grayscale image, wherein the input tumor contour is in the tumor contour annular region, and each of the plurality of reference segments is across the tumor contour annular region and includes a plurality of measured points; retrieving a modulated tumor contour suggestion point on each of the plurality of reference segments; and linking all the modulated tumor contour suggestion points on all of the reference segments, for forming the modulated tumor contour; wherein the modulated tumor contour suggestion point on each of the plurality of reference segments is obtained by the following steps: normalizing the value of all the measured points on the reference segment retrieved by a moving variance retrieving method, the value of all the measured points on the reference segment retrieved by a contrast retrieving method, the value of all the measured points on the reference segment retrieved by a distance retrieving method, the value of all the measured points on the reference segment retrieved by a gradient EWMA difference retrieving method, and the value of all the measured points on the reference segment retrieved by an angle retrieving method into value between 0 and 1; determining modulating criteria for each of the aforementioned retrieving methods; calculating a weighting parameter for each the measured points on the reference segment, by multiplying product corresponding to each of the aforementioned retrieving methods together, wherein each of the products is obtained by multiple the normalized value of the measured point for the number of times equal to the modulating criteria of the corresponding retrieving method, respectively; and retrieving the coordinate of the tumor contour suggestion point on the reference segment by calculating the average coordinate, from the sum of the product obtained by multiplying the coordinate of each of the measured points on the reference segment with the corresponding weighting parameter, respectively; wherein in the moving variance retrieving method, each of the plurality of reference segments includes a plurality of local segments and each of the plurality of local segments includes a plurality of moving horizons; each of the plurality of local segments consists of one of the plurality of measured points, at least one measured point before and at least one measured point after the one of the plurality of measured points, and each of the plurality of moving horizons consists of one of the plurality of measured points in the corresponding local segment and at least one measured point after the one of the plurality of measured points; and the value of all the measured points on the reference segment are retrieved by the moving variance retrieving method comprising the steps of: calculating a gradient variance of the grayscale image in the plurality of local segments and an average gradient variance of the grayscale image in the plurality of moving horizons; calculating the ratio of the gradient variance of the grayscale image in the plurality of local segments to the average gradient variance of the grayscale image in the plurality of moving horizons; and retrieving the measured point corresponding to the maximum ratio as the tumor contour suggestion point; wherein in the contrast retrieving method, a center of gravity of the input tumor contour is retrieved and a plurality of contrast reference segments is defined, each of the plurality of contrast reference segments extends from the center of the gravity to of the corresponding reference segments, respectively; and the value of all the measured points on the reference segment are retrieved by the contrast retrieving method comprising the steps of: calculating gradients of the grayscale image in the plurality of reference segments and an average gradient of the grayscale image in the plurality of contrast reference segments; calculating the differences between the gradients of the grayscale image in the plurality of reference segments and an average gradient of the grayscale image in the plurality of contrast reference segments; and retrieving the measured point corresponding to the maximum difference as the tumor contour suggestion point; wherein in the distance retrieving method, the value of all the measured points on the reference segment are retrieved in sequence by the distance retrieving method comprising the steps of: when there are no tumor contour suggestion points before, taking one of the plurality of measured points on the first reference segment as a first point, or when there is one tumor contour suggestion point before, taking a last tumor contour suggestion point on one of the plurality of reference segments as the first point; taking one of the plurality of measured points on the next reference segment as a second point; and retrieving the measured point on the next reference segment, having the minimum distance between the first point and the second point as the tumor contour suggestion point on the next reference point; wherein in the gradient EWMA difference retrieving method, the value of all the measured points on the reference segment are retrieved in sequence by the gradient EWMA difference retrieving method comprising the steps of: when there are no tumor contour suggestion points before, taking one of plurality of measured points on the first reference segment as a reference list; or when there is one tumor contour suggestion point before, taking all of the pervious tumor contour suggestion points as the reference list; providing an EWMA weighting formula; determining the weighting factor of the EWMA weighting formula and receiving the EWMA value of the reference list by the gradients of grayscale image of each of the measured points or the tumor contour suggestion points in the reference list; using the EWMA value of the reference list and the gradients of grayscale image of each of the plurality of measured points on the next reference segment as the input variables of the EWMA weighting formula; defining the output of the EWMA weighting formula as the EWMA value of each of the plurality of measured points on the next reference segment; and taking the one of plurality of measured points on the next reference segment, having the minimum difference between the gradient of grayscale image thereof and the EWMA value thereof, as the tumor contour suggestion point on the next reference segment; wherein in the angle retrieving method, the value of all the measured points on the reference segment are retrieved in sequence by the angle retrieving method comprising the steps of: when there are no tumor contour suggestion point before a last tumor contour suggestion point, defining one of the plurality of measured points on the first reference segment as a first point, and one of the plurality of measured points on the second reference segment as a second point; or when there is one tumor contour suggestion point before the last tumor contour suggestion point, defining the tumor contour suggestion point before the last tumor contour suggestion point as the first point, and defining the last tumor contour suggestion point as the second point; defining a vector going from the first point to the second point as a first vector, and vectors going from the second point to each of the plurality of measured points on a next reference segment as a plurality of second vectors; calculating the cosine values of the angles included by the first vector and the plurality of second vectors, respectively; and taking the one of plurality of measured points on the second next reference segment, being the end point of the second vector corresponding to the angle having maximum cosine value, as the tumor contour suggestion point on the second next reference segment.

3. A method of retrieving a tumor contour implemented in an image processing system that includes a memory storing a grayscale image, a display screen, an input device and a processor performing method steps, comprising: receiving an input tumor contour of the grayscale image; defining a tumor contour annular region and a plurality of reference segments of the grayscale image, wherein the input tumor contour is in the tumor contour annular region, and each of the plurality of reference segments is across the tumor contour annular region and includes a plurality of measured points; retrieving a tumor contour suggestion point on each of the plurality of reference segments; and linking all the tumor contour suggestion points on all of the reference segments, for forming the tumor contour, wherein a center of gravity of the input tumor contour is retrieved and a plurality of contrast reference segments is defined, each of the plurality of contrast reference segments extends from the center of the gravity to of the corresponding reference segments, respectively; and wherein the tumor contour suggestion point is retrieved by a contrast retrieving method comprising the steps of: calculating gradients of the grayscale image in the plurality of reference segments and an average gradient of the grayscale image in the plurality of contrast reference segments; calculating the differences between the gradients of the grayscale image in the plurality of reference segments and an average gradient of the grayscale image in the plurality of contrast reference segments; and retrieving the measured point corresponding to the maximum difference as the tumor contour suggestion point.

4. A method of retrieving a tumor contour implemented in an image processing system that includes a memory storing a grayscale image, a display screen, an input device and a processor performing method steps, comprising: receiving an input tumor contour of the grayscale image; defining a tumor contour annular region and a plurality of reference segments of the grayscale image, wherein the input tumor contour is in the tumor contour annular region, and each of the plurality of reference segments is across the tumor contour annular region and includes a plurality of measured points; retrieving a tumor contour suggestion point on each of the plurality of reference segments; and linking all the tumor contour suggestion points on all of the reference segments, for forming the tumor contour, wherein the tumor contour suggestion point is retrieved in sequence by a distance retrieving method comprising the steps of: when there are no tumor contour suggestion points before, taking one of the plurality of measured points on the first reference segment as a first point, or when there is one tumor contour suggestion point before, taking a last tumor contour suggestion point on one of the plurality of reference segments as the first point; taking one of the plurality of measured points on the next reference segment as a second point; and retrieving the measured point on the next reference segment, having the minimum distance between the first point and the second point as the tumor contour suggestion point on the next reference point.

5. A method of retrieving a tumor contour implemented in an image processing system that includes a memory storing a grayscale image, a display screen, an input device and a processor performing method steps, comprising: receiving an input tumor contour of the grayscale image; defining a tumor contour annular region and a plurality of reference segments of the grayscale image, wherein the input tumor contour is in the tumor contour annular region, and each of the plurality of reference segments is across the tumor contour annular region and includes a plurality of measured points; retrieving a tumor contour suggestion point on each of the plurality of reference segments; and linking all the tumor contour suggestion points on all of the reference segments, for forming the tumor contour, wherein the tumor contour suggestion point is retrieved in sequence by a gradient EWMA difference retrieving method comprising the steps of: when there are no tumor contour suggestion points before, taking one of plurality of measured points on the first reference segment as a reference list; or when there is one tumor contour suggestion point before, taking all of the pervious tumor contour suggestion points as the reference list; providing an EWMA weighting formula; determining the weighting factor of the EWMA weighting formula and receiving the EWMA value of the reference list by the gradients of grayscale image of each of the measured points or the tumor contour suggestion points in the reference list; using the EWMA value of the reference list and the gradients of grayscale image of each of the plurality of measured points on the next reference segment as the input variables of the EWMA weighting formula; defining the output of the EWMA weighting formula as the EWMA value of each of the plurality of measured points on the next reference segment; and taking the one of plurality of measured points on the next reference segment, having the minimum difference between the gradient of grayscale image thereof and the EWMA value thereof, as the tumor contour suggestion point on the next reference segment.

6. A method of retrieving a tumor contour implemented in an image processing system that includes a memory storing a grayscale image, a display screen, an input device and a processor performing method steps, comprising: receiving an input tumor contour of the grayscale image; defining a tumor contour annular region and a plurality of reference segments of the grayscale image, wherein the input tumor contour is in the tumor contour annular region, and each of the plurality of reference segments is across the tumor contour annular region and includes a plurality of measured points; retrieving a tumor contour suggestion point on each of the plurality of reference segments; and linking all the tumor contour suggestion points on all of the reference segments, for forming the tumor contour, wherein the tumor contour suggestion point is retrieved in sequence by an angle retrieving method comprising the steps of: when there are no tumor contour suggestion point before a last tumor contour suggestion point, defining one of the plurality of measured points on the first reference segment as a first point, and one of the plurality of measured points on the second reference segment as a second point; or when there is one tumor contour suggestion point before the last tumor contour suggestion point, defining the tumor contour suggestion point before the last tumor contour suggestion point as the first point, and defining the last tumor contour suggestion point as the second point; defining a vector going from the first point to the second point as a first vector, and vectors going from the second point to each of the plurality of measured points on a next reference segment as a plurality of second vectors; calculating the cosine values of the angles included by the first vector and the plurality of second vectors, respectively; and taking the one of plurality of measured points on the second next reference segment, being the end point of the second vector corresponding to the angle having maximum cosine value, as the tumor contour suggestion point on the second next reference segment.